Some possible theoretical variograms are shown in Figure 7 (from Barnes

et al., 1977). The horizontal axis represents distance h and the vertical
axis the value of the variogram y(h). The variogram in Figure 7(a) is
zero at h=0, increases, and then levels off.
The point where it levels
off, denoted h,, is the range; its interpretation is, that after distance

h,, the concentrations are no longer related (correlated). The value
yeh ) is called the “sill" and is the variance of the (independent)
variables greater than distance h, apart. Figure 7(b) represents a

variogram that increases proportional to the distance.
This implies
that the observations Z(x) and Z(x + h) are related no matter how large

the distance h.

This is hardly realistic in practice.

Figure 7(c)

illustrates a type of variogram that appears to be common in radionuclide
field studies; y(h) is discontinuous at 0.
A variogram must be 0 at h=0
since Z(x) - Z(x) = 0.
But Figure 7(c) shows a case where no matter how
close two points become, there is still some variability left.
The

discontinuity of the variogram at h=0 is called the "nugget effect."

Finally, Figure 7(d) represents a flat variogram, or pure nugget, meaning
the variables appear to be independent regardless of their proximity.
As we shall see, variograms of the soil and FIDLER data collected here

exhibit some of the features of types a, c, and d.
STUDY DESIGN AND RESULTS

The data in this study were taken along two transects, one near GZ and
the other 3,600 ft north of GZ at the Area 13 site.
The approximate
location of the two lines is given in Figure 8.
Note that Line l
(near GZ) crosses several activity level strata, while Line 2 is contained
in one of the lowest-level Am strata.
The data consists of both surface
soil samples analyzed for Am by Ge(Li) counting and FIDLER Am readings
in counts per minute.
For Line 1, three sets of FIDLER readings and
four adjacent ring samples (5-in. diameter and 5-cm deep) were taken
every 20 ft.
For Line 2, the interval was every 150 ft.
There were a
total of 15 such clusters of data for each line, spanning about 280 and

2,100 ft for Lines 1 and 2, respectively.

The three sets of FIDLER measurements consisted of

instrument placed directly on the ring,

(1) readings with the

(2) readings made with the

instrument held 1 ft above the ring, and (3) readings taken at the
center of the four samples at 1-ft height.
Two 70-g aliquots per ring
sample were counted using REECo's Ge(Li) system, giving a total of
120 soil concentrations per line.
At a later time,

100 adjacent FIDLER readings, both surface and 1-ft

height, were taken at 5-in. intervals 10 ft south of both lines.

The

purpose was to obtain information on the changes in activity occurring
on a scale of less than 40 ft.
The complete field sampling and laboratory
protocol for this study is given in Appendix C.

425

Select target paragraph3